The purpose of the proposed research is to test the effects of interventions that vary on locus of responsibility (LR) for self- care decisions (patient or health care provider) on the use of self- care behaviors and subsequent health outcomes among cancer patients receiving chemotherapy in the ambulatory care setting. The effects of an individual difference variable, desire for behavioral control of health care (BC), and various demographic variables will also be studied. Specific hypotheses about LR and BC effects on outcomes of coping (negative mood and disruption in day-to-day activities), morbidity, and utilization of self-care behaviors will be tested. The sample will consists of 162 patients beginning chemotherapy treatment on an outpatient basis. Patients will be randomly assigned to one of three groups, (i.e., usual care control, LR=patient or LR=professional responsibility). All subjects will receive a standardized informational intervention. The two experimental groups will listen to a tape recorded message focusing responsibility for decisions about self- care on either the patient or the professional and also will receive a follow-up telephone call from the nurse. Data on the outcomes of coping will be collected using the shortened Profile of Mood States and Selby's measure of disruption in day-to-day activity three times during each of the first two treatment cycles and again six months after the beginning of treatment. Side effect intensity and distress will be measured at the same times as the outcomes of coping. Data on use and efficacy of self-care behaviors will be collected during the second and ninth days of the first two treatment cycles. Coping strategies and use of general preventive strategies will be measure nine days after treatment for treatment cycles one and two. Repeated measures ANCOVA with subsequent contrast analyses will be used to test the hypotheses. Descriptive and correlational analysis will be used to identify the patterns of side effects and use/efficacy of self-care behaviors and to identify mediating variables that may explain the effects of the interventions.